Horizon-Labs
/

multilingual-embedding-base

Sentence Similarity

sentence-transformers

Transformers.js

feature-extraction

text-embeddings-inference

Model card Files Files and versions

Horizon-Labs commited on 25 days ago

Commit

02e8d8b

·

verified ·

1 Parent(s): 6d75a8f

Card: leaderboard link

Files changed (1) hide show

README.md +2 -1

README.md CHANGED Viewed

@@ -135,7 +135,8 @@ candidates each. "Other" = ESCI (es, jp, us), RuBQ, T2Reranking, mMARCO-ja and A
 - The table shows the released checkpoint. Means over two training seeds: small own .817 / against a bge-m3 index .818; base own .822 / against a bge-m3 index .816.
 - Models ahead of this one (own embeddings): MIRACL: bge-m3; Wikipedia: none;
   other: bge-m3.
-- This is a reranking-style test over given candidate lists, not a full-corpus retrieval benchmark.
 ## Training

 - The table shows the released checkpoint. Means over two training seeds: small own .817 / against a bge-m3 index .818; base own .822 / against a bge-m3 index .816.
 - Models ahead of this one (own embeddings): MIRACL: bge-m3; Wikipedia: none;
   other: bge-m3.
+- This is a reranking-style test over given candidate lists, not a full-corpus retrieval benchmark. Rerankers on the same test:
+  [leaderboard](https://huggingface.co/spaces/Horizon-Labs/multilingual-retrieval-leaderboard).
 ## Training